Nanoscience and Nanotechnology for Chemical ... - ACS Publications

Downloaded by UNIV OF DELAWARE MORRIS LIB on June 28, 2012 | http://pubs.acs.org Publication Date: December 24, 2009 | doi: 10.1021/bk-2009-1016.ch012

Chapter 12

Novel GaN-based Chemical Sensors for Long-range Chemical Threat Detection K.-A. Son1, B. H. Yang1, N. Prokopuk2, J. S. Moon3, A. Liao1, M. Gallegos1, J. W. Yang4, T. Katona4, and M. A. Khan4 1

Jet Propulsion Laboratory, California Institute of Technology, Pasadena,CA 2 Naval Air Warfare Center, China Lake, CA 3 HRL Laboratories, LLC, Malibu, CA 4 University of South Carolina, Columbia, SC

We are developing micro chemical sensor nodes consisting of GaN HEMT (High Electron Mobility Transistor) sensors and a RF communication link for long range chemical threat detection and early warning. In this paper, we discuss our research on (1) high selectivity detection of chemical agents (stimulants) using the GaN HEMT sensors, (2) the effects of the materials and design of the gate electrode on the sensitivity of the sensor, and (3) optimal operating parameters for high sensitivity detection.

Introduction Early warning capability of chemical or biological threats is crucial for the safety and security of the public as well as military forces, and it has been a driving force for development of stand-off detection technologies. The current stand-off monitoring systems for chemical or biological agents are mostly based on optical spectral analysis and deployed on mobile/fly-by platforms (1-3). These systems offer excellent chemical identification capability, but are in © 2009 American Chemical Society

155

In Nanoscience and Nanotechnology for Chemical and Biological Defense; Nagarajan, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2009.


156 general large, expensive, and require long data acquisition times. Functionality of the optical systems can be hampered by bad weather or by air clutter (e.g., smoke or fire) due to undesirable signal attenuation (1-3). For defense and security applications, real-time and autonomous monitoring of the continuously changing chemical and biological environment is highly beneficial. Continuous monitoring of the environment is also critical for proper implementation of the decontamination protocol. To this end, we are developing micro chemical sensor nodes that will enable real-time remote detection and monitoring of chemical threats and rapid transmission of the early warning signals. Our sensor node consists of an array of GaN-based micro chemical sensors, a micro controller, and a RF link. In this paper, we will discuss the selectivity and sensitivity of chemical agent (simulants) detection using GaN HEMT (High Electron Mobility Transistor) sensors. GaN HEMT devices have been developed mostly for RF (radio frequency) transceivers, and their potential for chemical or biological sensors has been recognized only recently. Unlike the ISFET (Ion-Sensitive FET), which is traditionally used for chemical sensing, the GaN HEMT has a two-dimensional electron gas (2DEG) conducting channel near the AlGaN/GaN interface caused by the differences in spontaneous polarizations between AlGaN and GaN layers and piezoelectric polarization of the pseudomorphic AlGaN layer (Fig. 1) (4-7). This 2DEG induces surface polarization sheet charges on the sensor surface. While the details of chemical sensing mechanism of GaN HEMT is not clearly known, it is generally understood that chemical species adsorbing on the GaN HEMT sensor can modify the surface polarization charges, modulating the charge density and conductivity of the 2DEG and thus providing a transduction mechanism for chemical sensing (8). In GaN HEMT, the AlGaN layer is typically 10-30 nm thick. Therefore the conduction within this 2DEG channel is very sensitive to surface polarization interactions induced by chemical species (on the sensor surface, especially on the gate electrode). The resulting transconductance change is amplified by the gain of the HEMT, offering potential for high sensitivity detection of chemical analytes with a very short response time of a few milliseconds. To date, GaN HEMT sensors have been investigated primarily for extreme environment sensors with focus on detecting automobile combustion gases (9) and hydrogen (10) and on measuring pressure or stress (11-13). Development toward biological sensors has also been reported recently for GaN HEMT sensors (14,15). In our work, GaN HEMT sensors have been studied for chemical agent detection with the ultimate goal of developing a chip-level GaN chemical sensor node (16). In this paper, we discuss our research on (1) high selectivity detection of chemical agents (stimulants) using the GaN HEMT sensors, (2) the effects of the materials and design of the gate electrode on the sensitivity of the sensor, and (3) optimal operating parameters for high sensitivity.


157

Experimental


The GaN HEMT sensors are fabricated with AlxGa1-xN/GaN (x=0.25-0.3) heterostructures grown on semi-insulating 4H-SiC substrates. The HEMT structures are grown by a combination of pulsed atomic layer epitaxy (PALE) and conventional metalorganic chemical vapor deposition (MOCVD). The AlGaN layer is 20-25 nm thick and with a thin AlN spacer layer above the 2DEG channel. The GaN layer was ~1300 nm thick and delta-doped. Typical sheet charge density is ~1x1013/cm2 for these structures, and electron mobility is ~ 1500 cm2/Vs. Mesa isolation of devices is carried out with chlorine-based reactive ion

Source

R 2 PO OF CN R 2P Gate

Drain Iout

AlGaN GaN

Source

2DEG

SiC Substrate

Gate

Drain

Source Gate 20 μm

Fig. 1 Schematic of a cross sectional view of a GaN HEMT sensor (left; not drawn to scale) and SEM image of a pair of GaN HEMT sensors (right). (Reproduced from reference 16. Copyright 2007 SPIE) etching. Ohmic contacts for the source and drain electrodes are made with Ti/Al/Ni/Au layers by rapid thermal annealing at 800°C, and specific contact resistance of ~0.5 ohm-mm is obtained. The excellent ohmic contact resistance results in a lower knee voltage, allowing the sensor operation at source-drain potential difference of less than 3 V. The GaN HEMT sensors are fabricated with various gate electrode designs; various gate length (L), gate width (W), and source-drain distance (D) to investigate potential correlation to sensor sensitivity. For characterization of sensor responses to chemical species, Ni- and Pt-based Schottky contacts are employed as gate electrodes. The GaN HEMT sensors were tested with DECNP (Diethyl cyanophosphonate, a Tabun simulant), sulfur hexafluoride (SF6, a nerve agent stimulant), and common chemical solvents including diethyl ether, dichloromethane, ethyl acetate, acetonitrile, acetone, and benzene. To measure sensor responses, drain current (Ids) of a GaN HEMT was monitored at a constant drain voltage (Vds) and a constant gate voltage (Vgs) during exposures to various gaseous environments.


158

Results and discussion

0.07

0.20

0.06

Current change (mA)

Current change (mA)


A prototype Ni-gated GaN HEMT sensor with L=0.15 μm, W=200 μm, and D=2 μm was used to investigate sensor responses to DECNP and common chemical solvents. Sensor evaluation was performed at Vds=1.0–2.0 V with Ids in the order of 10 mA, which corresponds to dissipating power of 75 mW/mm. For a background signal, Ids was measured first with a flow of nitrogen. A dilute sample of analyte was then injected to the test chamber while measuring Ids. For samples of low concentration analytes, the nitrogen gas saturated with an organic vapor was diluted with pure N2 using mass flow controllers.

0.05 0.04 0.03 0.02 0.01 0.00 -0.01

1200

1600

Time (s)

2000

0.15

O

O P CN O

10 ppm 30 s @ 25°C

0.10 0.05 0.00 -0.05 150

200

250

300

Time (sec)

Figure 2 Electrical responses of a prototype Ni-gated GaN HEMT sensor. Source-drain current (Ids) of the GaN HEMT sensor was measured during exposures to diethyl cyanophosphonate (DECNP), a Tabun simulant, at room temperature: (a) 90 s exposures to 0.1% DECNP in N2, and (b) a 30 s exposure to 10 ppm DECNP. Measurements were made with Vds of 1.5 V and Vgs of 0.0 V. The arrows in the plot (a) indicate the exposure initiation times. (Reproduced from reference 16. Copyright 2007 SPIE) Figure 2 shows sensor response to DECNP at room temperature. The left plot presents sensor’s response to repeated exposures of 0.1% DECNP in N2. Each exposure was 90 s long and the starting point of each exposure is marked with an arrow in the plot. The result indicates that sensor response is immediate and repeatable. As shown in the right plot, the prototype GaN HEMT sensor is capable of detecting 10 ppm DECNP at room temperature, resulting in a clear increase of Ids. The sensitivity of the GaN HEMT sensor can be further improved by optimization of AlGaN/GaN epitaxial structure and gate electrode design.


159

0.2 0.0075 % DECNP 0.0 0.2 0.1 % diethyl ether

Change in Current (mA)


0.0 0.2

0.1 % dichloromethane

0.0 0.2

0.1 % ethyl acetate

0.0 0.2

0.1 % acetonitrile

0.0 0.2 0.1 % acetone 0.0 0.2

0.1 % benzene

0.0 200

300

400

500

600

Time (s) Figure 3 Electrical response of a prototype Ni-gated GaN HEMT sensor to DECNP and common chemical solvents. Among all the chemical species tested, the sensor shows positive responses to only DECNP at the concentrations used for the testing. (Reproduced from reference 16. Copyright 2007 SPIE) To study selectivity of the sensor, the GaN HEMT has also been investigated with common chemical solvents. Figure 3 shows Ids response, i.e., Ids(analyte exposure)-Ids(background in nitrogen), measured during repeated exposure cycles to DECNP, diethyl ether, dichloromethane, ethyl acetate, acetonitrile, acetone, and benzene at room temperature. The concentrations of the organic vapors used for the test were 0.0075% for DECNP and 0.1% for all other organics. The results show that the GaN HEMT sensor generates positive responses to only DECNP at the given condition, despite its concentration being one order of magnitude lower than that of the other chemicals. Modification of the surface polarization interaction induced by phosphonate and cyano groups is presumed to be a major factor for the high selectivity observed for DECNP, but more detailed study is under progress for better understanding of the selectivity. Positive responses of the sensor have been observed for the other organics tested, except benzene, when the concentration was much higher (~10%). However, such high organic concentration is not relevant for our study since it is not a typical operating condition for the chemical agent sensors. No response to


160 benzene was expected based on the zero dipole moment of the molecule and the absence of a functional group that can induce strong polarization interactions on the sensor (gate electrode) surface at room temperature. 0.015 0.013

-1.5Vg

0.011

-2.0Vg

Ids (A)

0.009

-2.5Vg

0.007


0.005 0.003 0.001

-4.0Vg

-0.001 0.0

1.0

2.0

3.0

4.0

5.0

6.0

Vds (V)

Figure 4 Ids-Vds (source-drain current vs. source-drain voltage) curves of a GaN HEMT sensor measured during exposure to SF6 in a room ambient condition; solid lines. As a reference, Ids-Vds curves of the sensor measured without SF6 exposure in the same condition are also shown; dashed lines. The sensor was fabricated with a 2 μm long and 50 μm wide Ni gate electrode with a source-drain distance of 6 μm. Measurements were made at a gate voltage (Vgs) of -1.5, -2.0, -2.5 and -4.0 V. To probe correlation between sensor design and sensitivity, GaN HEMT sensors fabricated with Ni and Pt gate electrodes in dimensions of L=2, 4, 8, 12 μm, W= 5, 10, 25, 50 μm, D=W+4 μm have been studied. Sensor responses were measured in a room ambient condition with repeated exposures to a 25 sccm flow of pure SF6, which was directed to the sensor at a ~2.5 cm distance. SF6 was used for the test due to the fact that it is a stimulant of a nerve agent but it is non-toxic and stable in a room environment. While SF6 has zero dipole moment, it has low-lying and localized unoccupied electronic states and shows a positive electron affinity (rending it to form SF-6 readily) (17,18). Thus a decent amount of Ids increase resulting from electron transfer from the gate electrode can be expected with SF6. Figure 4 shows Ids-Vds (source-drain current vs. source-drain voltage) curves of a GaN HEMT sensor measured during exposure to SF6. Comparison with the curves measured without SF6 presents clear increase of Ids upon SF6 exposure, especially in the saturation region of the IdsVds. One thing to note in the figure is that the degree of Ids increase is dependent on the gate electrode voltage. This observation will be discussed in more detail in a later section.


161 0.10

(a) GaN HEMT w/ Ni gate

Δ Ids (mA)

0.08 0.06 0.04 0.02 0.00 60

90

120 150 180 210 240 270 300

(b) GaN HEMT w/ Pt gate

0.08 Δ Ids (mA)


30 0.10

0.06 0.04 0.02 0.00 30

60

90

120 150 180 210 240 270 300 Time (s)

Figure 5 Electrical response of GaN HEMT sensor to SF6 measured using the sensors fabricated with two different gate electrode metals: (a) Ni and (b) Pt. In both sensors, the gate electrode was 2μm long and 50 μm wide (L2W50), and the source-drain distance was 6 μm. The drain current (Ids) was monitored at the gate voltage (Vgs) of -3 V and the drain voltage (Vds) of 2 V. Each SF6 exposure was 15 s long with a duty cycle of 60 s for the Ni-gated sensor and 45 s for the Pt gated sensor. SF6 exposures were carried out in air using a 25 sccm pure SF6 flow directed from a ~2.5 cm distance. Figure 5 shows Ids change measured with both Ni- and Pt-gated GaN HEMT sensors during exposure to SF6. In both cases, the sensors show immediate and repeatable Ids increases upon SF6 exposures. Based on the similarity of response observed with Ni- and Pt-gated sensors, in the following sections we present only the data acquired with Pt-gated GaN HEMT sensors.


162 0.045

0.0 Vgs

0.040 0.035

Ids (A)

0.030 0.025 0.020 0.015 0.010 0.005

-4.5 Vgs

0.000 -0.005

Ids (A)

0.045 0.040 0.035 0.030 0.025 0.020 0.015 0.010 0.005 0.000 -5.0

1.0

2.0

3.0

4.0

5.0

Vds (V)

6.0

5.5 Vds

0.5 Vds

-4.0

-3.0

-2.0

-1.0

0.0

Vgs (V)

0.014 0.012 dIds/dVgs (S)


0.0

0.010

5.5 Vds

0.008 0.006 0.004 0.002 0.000 -5.0

0.5 Vds -4.0

-3.0

-2.0

-1.0

0.0

Vgs (V)

Figure 6 Electrical characteristics of the GaN HEMT sensor used for testing with SF6. The sensor was fabricated with a 2 μm long and 50 μm wide (L2W50) Pt gate electrode with a source-drain distance of 6 μm. (a) Ids-Vds curves measured at the gate bias Vgs of -4.5 V (bottom) to 0.0 V (top); the curves for-4.5Vgs and -4.0Vgs are overlapping at Ids~0.0 mA. (b) Ids-Vgs transfer curves measured at the drain bias Vds of 0.5 V (bottom) to 5.5 V (top) (c) transconductance (dIds/dVgs vs. Vgs) of the GaN HEMT sensor measured at the drain voltage Vds of 0.5 V (bottom) to 5.5 V (top). Prior to probing factors governing sensor sensitivity, we analyzed electrical characteristics of the GaN HEMT sensors first. Figure 6 presents Ids-Vds (sourcedrain current vs. source-drain voltage), Ids-Vgs (source-drain current vs. sourcegate voltage) transfer characteristics, and transconductance (dIds/dVgs vs. Vgs) characteristics of the GaN HEMT sensor used for testing with SF6. The gate


163

The sensor resistance in the linear region of the Ids-Vds curve is 2.5 ohmmm, indicating that the ohmic contact resistance is quite small. The ohmic contact resistivity of the sensor that we measured is ~3.5x10-6 ohm cm2. As shown in Figure 4, upon exposure to the analyte gases, the sensors show a greater response in the saturation regime of the Ids-Vgs curves than in the linear regime. Clear understanding on the mechanism behind this observation requires theoretical modeling, which will be discussed elsewhere. Responses (Ids-time curves) of the GaN HEMT sensor shown in this paper were measured in the saturation regime. The Ids-Vgs transfer curves and the transconductance characteristics of the GaN HEMT sensor shown in the figure were measured at the Vds of 0.5 to 5.5 V. The peak transconductance is 240 mS/mm at Vds=5 V, indicating that the GaN HEMT sensor has excellent transconductance. In short, the GaN HEMT sensors are fabricated with excellent ohmic contact resistivity and small knee voltages (i.e., Vds at the start of Ids saturation), enabling sensor operation at Vds< 3 V and consequently with a small DC power dissipation.

0.115 0.095

Δ Ids (mA)


electrode of the sensor is a 2 μm long and 50 μm wide (L2W50) and 40 nm thick Pt-based Schottky contact located in the middle of the 6 μm gap between the source and the drain electrode. The Ids-Vds curves were measured at the Vg of -4.5 V to 0.0 V.

0.075

(a)

0.055

(b)

0.035

(c)

0.015

(d)

L2W50 L2W25 L2W10 L2W05

-0.005 30

90

150

210

270

Time (s)

Figure 7 Effects of gate width on the responses of GaN HEMT sensors to SF6. The GaN sensors were fabricated using Pt gate electrodes with various gate widths (W=5, 10, 25, & 50 μm) at a fixed gate length (L=2 μm). Source-drain current (Ids) of the GaN sensors was monitored at -3Vgs and 2Vds during exposures to SF6 at room temperature. Each SF6 exposure was 15 s long with a duty cycle of 45 s. The curves shown above correspond to Δ Ids.= Ids.(measured with SF6 exposure cycles)- Ids.(background measured with no SF6). For easier comparison between the curves, an offset value was added to each curve.


164 Effects of a sensor layout on detection sensitivity are investigated with the GaN HEMT sensors fabricated with various gate lengths (L) and gate widths (W). Figure 7 shows Ids response to SF6 exposures measured using the GaN HEMT sensors fabricated with W=5, 10, 25, 50 μm at L=2 μm. Higher sensitivity with an increasing gate width is clear. Figure 8 shows Ids response measured using GaN HEMT sensors fabricated with L=2, 4, 8 μm at W=50 μm and with L=2 μm at W=25 μm; in these sensors the source-drain distance (D) is D=L+4 μm.

L4W50

0.06 ΔIds (mA)

Δ Ids (mA)

0.07

L2W50

0.04 0.03 0.02 0.01 0.00

0.05 0.04 0.03 0.02 0.01 0.00

30

90

150

210

30

270

90

150

210

270

Time (s)

Time (s)

0.06

0.08 0.07

0.05

0.06

L8W50

0.04

Δ Ids (mA)

ΔIds (mA)


0.08

0.08 0.07 0.06 0.05

0.03 0.02

L2W25

0.05 0.04 0.03 0.02

0.01

0.01 0.00

0.00 30

90

150 Time (s)

210

270

30

90

150

210

270

Time (s)

Figure 8 Responses of GaN HEMT sensors to SF6 measured with various gate electrode designs: (a) L2W50, (b) L4W40 (c) L8W50 (d) L2W25, where L is the length of a gate electrode and W is the width of the gate electrode. The GaN sensors were fabricated using Pt gate electrodes, and the sourcedrain current (Ids) of the GaN sensors was monitored at -3Vgs and 2Vds during exposures to SF6 at room temperature. Each SF6 exposure was 15 s long with a duty cycle of 45 s. The curves shown above correspond to Δ Ids.= Ids.(measured with SF6 exposure cycles)- Ids.(background measured with no SF6). SF6 exposures were made in air using a 25 sccm pure SF6 flow directed from a ~2.5 cm distance. The data clearly show decreasing sensitivity with an increasing gate length at W=50 μm. Comparison between L4W50, L8W50, and L2W25 sensors, which corresponds to D8W50, D12W50, and D6W25, respectively, indicates that the sensor sensitivity is not simply proportional to Ids. The Ids of the GaN HEMT is proportional to the gate width W, and the L4W50 and the L8W50 sensors should have higher Ids than the L2W25 sensor. However, the L2W25 sensor shows a higher sensitivity, indicting that the shorter gate length plays a significant role.


165


The results shown in Figure 7 and 8 indicate that the sensor fabricated with L2W50 show excellent response compared to the other layouts and that the sensor response is strongly dependent on the gate length. These phenomena can be attributed to the fact that the sensor with L2W50 has the highest electric field near the gate-to-drain edge (in the saturation regime of the Ids-Vds), which is modulated by the adsorbed SF6 on the surface. The results also suggest that sensor sensitivity is not simply proportional to the size of the gate electrode or the amount of Isd of the sensor and that a short gate length and thus a sourcedrain distance are important factors in determining the sensitivity of the sensor. In order to find optimal operational parameters for GaN HEMT sensors, sensitivity dependence on the gate electrode voltage (Vg) has been studied.


166 Transconductance of GaN CWA sensor @ 2.0Vds

0.014

dIds/dVgs (S)

0.012 0.010

Ids change during exposures to SF6

0.20

0.15

0.008 0.10

0.006 0.004

0.05

-2.0Vgs

0.002 0.000

0.00

-5.0

-4.0

0.20

-3.0

-2.0

-1.0

0.0

50

Δ Ids (mA)

0.20

Δ Ids (mA)

250

300

-1.5Vgs

-4.0Vgs 0.05

0.00

0.00 50

100

150

200

250

300

0.20

50

100

150

200

250

300

Time (s)

0.15

0.15

0.10

0.10

-3.5Vgs

0.05

0.00

-1.0Vgs

0.05

0.00

0.20 50

Δ Ids (mA)

200

Time (s)

0.10

0.10

0.05

100

150

200

250

300

0.20

0.15

0.15

0.10

0.10

-3.0Vgs

0.05

0.00

50

100

150

200

250

300

-0.5Vgs 0.05

0.00

50

Δ Ids (mA)

150

0.15

0.15


100

0.20

Vgs (V)

100

150

200

250

300

0.20

0.20

0.15

0.15

0.10

50

100

150

200

250

300

0.10

-2.5Vgs

0.05

0.00

0.0Vgs

0.05

0.00

50

100

150

200

Time (s)

250

300

50

100

150

200

250

300

Time (s)

Figure 9 Transfer characteristics of a GaN HEMT sensor fabricated with a L2W50 Pt gate electrode (top left) and Ids responses of the sensor to SF6 measured at various gate electrode voltages(Vgs=-4.0 to 0.0 V). For all the measurements shown above, the drain voltage(Vgs) was 2.0 V. Figure 9 shows Ids responses of a L2W50 sensor to SF6 exposures measured at various gate voltages Vg at a constant Vds=2.0V. Transconductance curve of the L2W50 sensor measured at Vds=2.0V is also shown in the figure as a reference. The data show that the highest sensitivity or the peak response is


167 obtained when the gate voltage is applied near the peak transconductance. The data indicate a strong correlation between sensor sensitivity and transconductance; sensitivity increases with increasing transconductance. Once the transconductance passes the maximum, the sensitivity and the signal to noise ratio (S/N) seem to drop rapidly. Since the device transconductance is dependent on the gate capacitance, this observation can be attributed to the surface charge accumulated due to the exposure to SF6 or modulation of gate voltage Vg induced by adsorbing SF6 molecules.


Summary We are developing GaN HEMT-based micro chemical sensors and sensor nodes for remotely operated long-range detection of chemical threats. In this paper, our GaN HEMT micro chemical sensor study is described with various device design layouts and wide range of operational parameters using DECNP and SF6, chemical agent simulants. High sensitivity (10 ppm) and high selectivity detection of DECNP has been demonstrated at room temperature using a prototype Ni-gated GaN HEMT sensor fabricated with L=0.15 μm, W=200 μm, and D=2 μm. The sensitivity study performed with various GaN HEMT sensors (L=2, 4, 8, 12 μm, W= 5, 10, 25, 50 μm,) using SF6 indicates that GaN HEMT sensor with a shorter gate length and a larger gate width exhibit a higher sensitivity. The GaN sensors also show higher sensitivity in the saturation region of Ids-Vds compared to the linear region. Furthermore, the sensor sensitivity is strongly dependent on transconductance of the sensor; the highest sensitivities are measured near the gate voltage of the peak transconductance. Based on the current observations with SF6, it is expected that further scaling of the gate-to-channel (2DEG) distance and of the gate length will lead to higher sensitivity detection of chemical agents (simulants) using GaN HEMT sensors.

Acknowledgements The research described in this paper was carried out by the Jet Propulsion Laboratory, California Institute of Technology, and was sponsored by ARO/DTRA and monitored by Dr. Stephen J. Lee at ARO. The authors thank Dr. C. J. Whitchurch for valuable input for the project and Prof. P. P. Ruden for helpful discussions.


168

References 1. 2.


3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Fountain, A. W. III Ed. Chemical and Biological Sensing VIII, Proceedings of SPIE, Vol. 6554 (2007). Ewing, K. J.; Gillespie, J. B.; Chu, P. M.; Marinelli, W. J. Eds. Chemical and Biological Sensors for Industrial and Environment Monitoring III, Proceedings of SPIE, Vol. 6756 (2007). Jensen, J. O. ; Theriault, J. Eds. Chemical and Biological Standoff Detection III, Proceedings of SPIE, Vol. 5995 (2005) Bernardini, F.; Fiorentini, V.; Vanderbilt, D. Phys. Rev. B, 1997, 56, R10024-R10027. Ambacher, O.; Smart, J.; Shealy, J. R.; Weimann, N. G.; Chu, K. et al. J. Appl. Phys. 1999, 85, 3222-3233 Ambacher, O.; Foutz, B.; Smart, J.; Shealy, J. R.; Weimann, N.G. et al. J. Appl. Phys. 2000, 87, 334-344. Yu, E. T.; Dang, X. Z.; Asbeck, P. M.; Lau, S. S.; Sullivan, G. J. J. Vac. Sci. & Tech. 1999, 17, 1742-1749 Schlwig, J.; Muller, G.; Eickhoff, M.; Ambacher, O.; Stutzmann, M. Sensors and Actuators 2002, B87, 425-430. Schlwig, J.; Muller, G.; Eickhoff, M.; Ambacher, O.; Stutzmann, M. Mat. Sci. Eng. 2002, B93, 207-214 Wang, H.; Kang, B. S.; Ren, F.; Fitch, R. C.; Gillespie, J. K. et al. Appl. Phys. Lett. 2005, 87, 172105. Liu, Y. ; Ruden, P.P.; Xie, J.; Morkoç, H.; Son, K.-A. Appl. Phys. Lett. 2006, 88, 013505. Son, K. -A.; Yang, B. H.; Liao, A.; Steinke, I. P.; Liu, Y.; Ruden, et al. proceedings of NASA Science and Technology Conference, 2007. Kang, B. S.; Kim, J. Jang, S.; Ren, F.; Johnson, J. W. et al. Appl. Phys. Lett. 2005, 86, 253502. Kang, B. S.; Ren, F.; Wang, L; Lofton, C.; Tan, W. W. et al. Appl. Phys. Lett. 2005, 87, 0253508. Steinhoff, G.; Baur, B.; Wrobel, G.; Ingebrandt, S.; Offenhausser, A.; Dadgar A. et al. Appl. Phys. Lett. 2005, 86, 033901. Son, K.-A.; Yang, B.; Prokopuk, N.; Moon, J.S.; Liao, A.; Gallegos, M.; Yang, J. W.; Khan, M.A. Proceedings of SPIE, 2007, vol.6556 Micro (MEMS) and Nanotechnologies for Defense and Security, paper # 6556-36 Klekamp, A.; Umbach, E. Surf. Sci. 1992, 271, 555-566. Faradzhev, N. S.; Kusmierek, D. O.; Yakshinskiy, B. Y.; Madey, T.E. Low Temp. Phys. 2003, 29, 215-222.


Nanoscience and Nanotechnology for Chemical ... - ACS Publications

Recommend Documents